Many dog-owning homeowners and renters find it challenging to acquire liability insurance coverage if their dogs are a breed considered by an insurer to pose a higher risk than other dogs. These dog owners are often denied coverage, have policies canceled or not renewed, or pay higher premiums based merely on the dog’s breed (or a mix of breeds) with no consideration given to the actual risk posed by a specific dog’s behavior. The resulting unequal treatment between dog owners raises the concern that breed-based limitations are unfairly discriminatory and violate state insurance laws.
Some states have addressed this issue, acknowledging that making distinctions between dog breeds is a valid concern. They have sought more data from insurance companies to learn about the risk or banned the practice altogether. Recently, the National Council of Insurance Legislators (NCOIL) took up the issue and, in collaboration with advocates on both sides of the issue, approved the Dog Breed Insurance Underwriting Protection Act, a model law drafted to prohibit the use of dog breed as a factor in homeowners’ and renters’ insurance decisions.
Dog Breed Lists
Many, but not all, insurance companies have created lists of dog breeds (e.g., German shepherds, Rottweilers, pit bulls) and mixes of these breeds that they consider uninsurable based on assumptions about a breed’s intrinsic behavior. The problem, though, is that the assumptions are unsupportable. No reliable actuarial data justifies making risk distinctions between dog breeds.
Initial dog breed lists were based on outdated studies conducted by the Center for Disease Control (CDC). However, the studies are unreliable because of limitations in accurate data gathering and the inability to identify dog breeds accurately. Indeed, recognizing these limitations, the CDC no longer gathers dog bite data and has distanced itself from the original studies by noting that aggression in dogs is influenced by numerous factors unrelated to breed, including the dog’s sex, health, socialization, and even the victim’s behavior. Subsequent studies by other researchers fall on both sides of the debate, with some suggesting breed risk and others disputing it. Each of the studies is plagued by similar limitations—data collection methods that rely on inconsistent breed identification, a disproportionate focus on dog-bite fatalities, and a lack of a baseline for dog breeds in the United States, making it impossible to identify risk accurately.
State law generally prohibits insurers from treating individuals or risks of the same class and hazard differently unless the different treatment is supported by sound underwriting principles and actuarial data related to the anticipated loss. Thus, if an insurer plans to charge one premium to Homeowner A, who owns a beagle, but a higher premium to Homeowner B, who is the same in all respects except that the dog is a Rottweiler, the insurer must have reliable actuarial data to support the different premiums. Without that data, there is no justification for treating the homeowners differently. And that is the problem—a lack of reliable actuarial data to support using dog breed to differentiate risk between homeowners.
Concern over this lack of data has compelled advocates to seek equity for homeowners from legislators and insurance regulators, and states have responded with either desk drawer rules or legislation that address the matter in two ways. Some states, like Nevada’s NRS § 687B.383 and New York’s INS § 3421, prohibit outright different treatment of dog owners based solely on dog breeds. However, both states make clear that the prohibition does not apply if the homeowner’s dog has been designated “dangerous” or “vicious” under state statute. Nevada has taken even a step further by prohibiting insurers from asking about a dog’s specific breed or mixture. However, they may ask whether the dog has been declared dangerous or vicious. And Arizona’s ARS § 11-1025, which has similar prohibitions, also proscribes consideration of dog breed when deciding whether a dog is “aggressive” or “vicious” or “has created liability.”
Other states have taken a different approach, mandating a data call for more information about dog-related insurance claims. Illinois, for example, imposed a two-year data call beginning this year. Under § 143.10d of the Illinois Insurance Code, insurers offering homeowners or renters insurance must provide the Illinois Department of Insurance data on each dog-related claim received, including information on the breed of dog at issue, who identified the breed if identified visually, from whom the owner acquired the dog (e.g., pet store, breeder, shelter), sex, spay/neuter status, and training of the dog, whether the dog was provoked, the type of injury sustained, and whether there were prior claims or complaints about the dog. Massachusetts issued a three-year data call that concluded this year seeking even more detailed information, including not only the above but also data on the particulars of the incident, the injured party, control over the dog, and the incident’s location.
NCOIL Model Law
NCOIL’s adoption of a model law on breed discrimination serves two purposes. First, the discussion educated state insurance legislators on the inequities posed by using dog breed lists. Second, the law encourages greater consistency between states when legislating on the issue.
The model law takes the direct prohibition approach, providing in pertinent parts:
(A) With respect to homeowners’ insurance policies and renter’s insurance policies . . . , no insurer shall refuse to issue or renew, cancel, or charge or impose an increased premium or rate for such policy or contract, or exclude, limit, restrict, or reduce coverage under such policy or contract based solely upon harboring or owning any dog of a specific breed or mixture of breeds.
(B) The provisions of this section shall not prohibit an insurer from refusing to issue or renew or from canceling any such contract or policy, nor from imposing a reasonably increased premium or rate for such a policy or contract based upon the designation of a dog of any breed or mixture of breeds as a dangerous dog . . . , based on sound underwriting and actuarial principles reasonably related to actual or anticipated loss experience . . . .
The model law follows the language of New York’s law but also prohibits basing exclusions, limitations, restrictions, or reductions in coverage on breed alone. That the model law takes the prohibition approach over the data call approach indicates that NCOIL considers breed-based distinctions unsupportable by actuarial data. However, the law clarifies that the prohibited conduct uses dog breeds as the sole consideration for treating a homeowner or renter differently than others. This is further explained in subsection (B), which provides that a dangerous dog designation can be a supportable basis for distinction.
What does this mean going forward? Insurers that rely on dog breed lists in making policy determinations should be aware that the practice may be on its way out. Although states are not required to adopt the model code, the fact that insurance legislators are now aware of the issue makes legislation more likely. And for consumers who own listed dog breeds, insurance coverage may become easier to find and more affordable if insurers anticipate legislation and choose to end the use of breed lists voluntarily.